As most anyone who has recently applied for a job knows, hiring has changed dramatically in recent years. The Internet has replaced job advertisements in newspapers, one of the key factors driving the financial decline of the latter, and software has replaced most recruiters. Because job applications are done online, applicants rarely talk to anyone, even by e-mail, during the hiring process.
One upside of this automation is that applying for jobs has been made considerably easier, an outcome that was intended in the 1990s, when these systems were born and employers were competing to attract applicants. But there has been an unintended downside: That ease, combined with the huge pool of job seekers, now means that employers are overwhelmed with job applications. At the same time, human resource departments have been pushed to cut costs, especially their own head count. The only way to meet those two demands has been to move even further toward automating the entire hiring process.
(MORE: How to Buy a House)
Once job requirements have been settled on, however haphazardly, they are then built into the hiring software that screens applications. Some requirements, such as possessing credentials, are easy to write into software, but others, like the ability to get along with customers, are not, and may require many questions even to get close to an accurate response. And even then, the software might be identifying the wrong qualities actually needed for the vacant position. Tom Keebler at the HR consulting firm Towers Watson, who advises employers about hiring systems, says even well-intentioned hiring managers have a problem trying to identify skills that are not easily associated with credentials or experience.
Once it’s in the software, each requirement, critical or trivial, essentially becomes something like a hurdle that applicants have to clear to become a qualified candidate. One job seeker who wished not to be identified described this experience with an employer: “I asked if anyone had done the job yet successfully with the additional duties and they said, ‘Yes, the lady who is temping in that position is currently doing all of those duties and she does them well.’ So I asked why they didn’t hire her and they told me that she failed the online questionnaire (which was mostly about personality fit) so they wouldn’t hire her.”
Apparently doing the job well wasn’t enough of a qualification. This was, it should be noted, no isolated tale. Thirty-eight percent of employees in a 2009 Business Roundtable survey reported that their current employer looks only at education and prior experience, rather than directly at skills and abilities, to determine who can do which job.
The criterion placed in these software programs is having actually done the tasks before, not just being able to do them, but woe to the applicant, however able, whose experience and credentials don’t form a perfect match with what the software is looking for. Jeffrey Oleander (his nom de Net) says, “I once designed and developed a set of tools for software quality testing management but was turned down for [a job] that used a particular brand-name version of just such a tool.” Having built such systems wasn’t enough, either. Says Oleander, “Another time, I was turned down because I didn’t have 2 years of experience using an extremely simple database report formatting tool, the sort of thing that would require just a couple hours for any half-decent database wrangler to master, less than an hour for the very best.”
(MORE: Your Global Economic Mess is Now Being Served)
Then there is the wage issue. Most of these automated systems ask the applicant about the wage he or she thinks is acceptable. Some systems tell the applicants the wage and ask if that is okay. If respondents say no, their application is put aside. Maybe that’s reasonable if the employer can’t pay more, but labeling such a candidate as unqualified tortures the meaning of “applicant shortage.”
In effect, leaving wage questions up to anonymous software creates an auction approach where potential applicants feel as though they are bidding for the job. It’s like the old limbo dance challenge: How low can you go? Applicants might win jobs by underbidding the competition, but they are apt to saddle themselves with a wage well below the market rate. Yet if they guess too high, they can be assured that the software will kick their applications out of the running. Damned if they do, damned if they don’t.
How sophisticated are the algorithms used to parse job applications for the right attributes? Mightily so, but they are capricious, too. This Beat-the-Software advice, culled from various experts, should serve as a warning shot across the bow of employers and would-be employers equally. Does anyone benefit from a hiring process that turns on such small distinctions?
- Don’t use headers or footers. They jam most parsing algorithms.
- Customize each résumé based on language used in the job description. If the description says “CPA,” make sure “CPA” is on your résumé. Don’t go too far, though: copying and pasting the job description won’t land you the gig.
- Use conventional formats. While fancy fonts, strange layouts, and functional formatting might impress an employer, computers hate them. Stick to a simplistic style and reverse chronological formatting.
- Put it in context. Modern résumé parsers check the context of buzzwords such as Java or C++, so if you want to seem different from the kid who took one “Java” class in high school, go more in depth about what you know and how long you’ve known it.
- Submit your résumé in text format. While .pdf might be convenient, MS Word generally ensures the least parsing errors.
- Never use graphics. Besides being unprofessional, graphics always hamper the parsing process and generally show up as white noise to the algorithm. White noise is just what you don’t want.
- Include your postal address. Your address is often how your résumé is filed. If you don’t include it, you might not get considered at all.
Peter Cappelli is Professor of Management and Director of Wharton’s Center for Human Resources. He was previously co-director of the National Center on the Educational Quality of the Workforce for the U.S. Department of Education. This article is the second of two adapted from his new book, Why Good People Can’t Get Jobs: The Skills Gap and What Companies Can Do About It, which was published earlier this month. The first article can be found here.